Large Reasoning Models (LRMs) have advanced automated multi-step reasoning, but their ability to generate complex Chain-of-Thought (CoT) trajectories introduces severe privacy risks, as sensitive information may be deeply embedded throughout the reasoning process. Existing Large Language Models (LLMs) unlearning approaches that typically focus on modifying only final answers are insufficient for LRMs, as they fail to remove sensitive content from intermediate steps, leading to persistent privacy leakage and degraded security. To address these challenges, we propose Sensitive Trajectory Regulation (STaR), a parameter-free, inference-time unlearning framework that achieves robust privacy protection throughout the reasoning process. Specifically, we first identify sensitive content via semantic-aware detection. Then, we inject global safety constraints through secure prompt prefix. Next, we perform trajectory-aware suppression to dynamically block sensitive content across the entire reasoning chain. Finally, we apply token-level adaptive filtering to prevent both exact and paraphrased sensitive tokens during generation. Furthermore, to overcome the inadequacies of existing evaluation protocols, we introduce two metrics: Multi-Decoding Consistency Assessment (MCS), which measures the consistency of unlearning across diverse decoding strategies, and Multi-Granularity Membership Inference Attack (MIA) Evaluation, which quantifies privacy protection at both answer and reasoning-chain levels. Experiments on the R-TOFU benchmark demonstrate that STaR achieves comprehensive and stable unlearning with minimal utility loss, setting a new standard for privacy-preserving reasoning in LRMs.
LMU Munich(慕尼黑大学);Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心);Technical University of Munich(慕尼黑技术大学);Siemens AG(西门子股份公司);University of Oxford(牛津大学)
CURE:Circuit-Aware Unlearning for LLM-based Recommendation
University of Massachusetts Lowell(马萨诸塞大学洛厄尔分校);Zhejiang University(浙江大学);The Chinese University of Hong Kong(香港中文大学);National University of Singapore(新加坡国立大学)
Forgetting-MarI: LLM Unlearning via Marginal Information Regularization
Department of Mathematics, University of California Davis, USA(加州大学戴维斯分校数学系);SLAC National Accelerator Laboratory, Stanford University, USA(斯坦福大学SLAC国家加速器实验室);Department of Computer Science, University of California Davis, USA(加州大学戴维斯分校计算机科学系)
Step-by-Step Reasoning Attack: Revealing 'Erased' Knowledge in Large Language Models
School of Computing(计算机学院);National University of Singapore(新加坡国立大学);Department of Electrical and Computer Engineering(电气与计算机工程系);College of Design and Engineering(设计与工程学院);RespAI Lab(RespAI实验室);School of Computer Engineering(计算机工程学院);Institute for Infocomm Research(信息与通信研究所);A*STAR
Breaking the Trilemma of Privacy, Utility, Efficiency via Controllable Machine Unlearning
University of Notre Dame(圣母大学);University of Pennsylvania(宾夕法尼亚大学);Dartmouth College(达特茅斯学院);Georgia Institute of Technology(佐治亚理工学院);George Mason University(乔治梅森大学)
A Survey on Recommendation Unlearning: Fundamentals, Taxonomy, Evaluation, and Open Questions
School of Communication Engineering, Hangzhou Dianzi University(杭州电子科技大学通信工程学院);College of Computer Science and Technology, Zhejiang University(浙江大学计算机科学与技术学院);Hong Kong Polytechnic University, Academy for AI(香港理工大学人工智能学院)
Machine Unlearning of Traffic State Estimation and Prediction
Department of Civil and Environmental Engineering, University of Washington(土木与环境工程系,华盛顿大学);Department of Mathematics, University of Washington(数学系,华盛顿大学);(Jeff) Ban(Ban)
Forget and Explain: Transparent Verification of GNN Unlearning
Chung-Ang University Department of Smart Cities Seoul Republic of Korea(Chung-Ang大学智慧城市系 首尔 大韩民国);Chung-Ang University Department of Computer Science(Chung-Ang大学计算机科学系);Chung-Ang University(Chung-Ang大学)